1 Set Up

1.1 Load Libraries

1.2 Experiment Parameters

1.2.1 Set condition names relevant to the RSA

1.2.2 Define Hypothesized ROIs

1.2.3 Define color palette for plots

2 Positive and Negative Outcome Similarity

In this analysis, we will directly test whether an ROI represents positive and negative wins in a similar manner. This will be done by correlating the multi-voxel representation of positive win trials to negative win trials. The hypothesis here is that the posterior cerebellum will respond similarily when participants were correct about someone liking them as being correct about someone disliking them, and that this similarity will be greater than in the ventral striatum.

2.1 Set Up

2.1.1 Load data

This analysis was completed using the Nilearn python package. Here we will only conduct second level ROI statistics and visualizations.

Divide data by task

Filter data by hypothesized ROIs

Fischer r to z transform (this is not used)

2.2 Conduct One Sample T-Tests

These t-tests will examine whether each ROI had significantly greater similarity than 0

2.3 Monetary

2.3.1 ANOVA

##              Df Sum Sq Mean Sq F value  Pr(>F)   
## roi           2 0.0640 0.03202   6.168 0.00256 **
## Residuals   180 0.9344 0.00519                   
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##       region02   region08 striatum_ventral
## mean 0.0612427 0.06177925       0.02183083
## sd   0.0635708 0.09355805       0.05270933

2.3.2 Post-hoc Tests

##   Tukey multiple comparisons of means
##     95% family-wise confidence level
## 
## Fit: aov(formula = r ~ roi, data = all_sim_vcr_mdoors_rois)
## 
## $roi
##                                    diff         lwr          upr     p adj
## region08-region02          0.0005365529 -0.03029410  0.031367202 0.9990678
## striatum_ventral-region02 -0.0394118682 -0.07024252 -0.008581219 0.0080823
## striatum_ventral-region08 -0.0399484212 -0.07077907 -0.009117772 0.0071204

2.3.3 Visualization

Specify ROI plotting order on the x-axis

2.4 Social

2.4.1 ANOVA

##              Df Sum Sq Mean Sq F value Pr(>F)    
## roi           2 0.2854 0.14269   18.08  7e-08 ***
## Residuals   180 1.4207 0.00789                   
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##        region02  region08 striatum_ventral
## mean 0.07467777 0.1176262       0.02109065
## sd   0.06479725 0.1320638       0.04516011
##         region02         region08 striatum_ventral 
##       0.06479725       0.13206380       0.04516011

2.4.2 Post-hoc Tests

##   Tukey multiple comparisons of means
##     95% family-wise confidence level
## 
## Fit: aov(formula = r ~ roi, data = all_sim_vcr_social_rois)
## 
## $roi
##                                  diff         lwr         upr     p adj
## region08-region02          0.04294842  0.00493096  0.08096587 0.0224537
## striatum_ventral-region02 -0.05358712 -0.09160458 -0.01556966 0.0030011
## striatum_ventral-region08 -0.09653554 -0.13455299 -0.05851808 0.0000000

2.4.3 Visualization

3 Theoretical Representational Similarity Analysis

In this analysis we will create theorical RDMs, that predict a specific representation within an ROI.

3.1 Theoretical RDMs

3.1.1 Task Domain

3.1.2 Valence

3.1.3 Outcome

3.1.4 Outcome Valence

Being correct about someone not liking you is different from being correct about someone liking you. But being incorrect about someone not liking you is the same feeling as being incorrect about someone not liking out (negative).

3.1.5 Interaction

Being correct about someone not liking you is different from being correct about someone liking you. But being incorrect about someone not liking you is the same feeling as being incorrect about someone not liking out (negative).

3.2 Task Domain

Filter data by hypothesized ROIs

3.2.1 Conduct one sample t-tests

3.2.2 ANOVA

##              Df Sum Sq Mean Sq F value Pr(>F)
## roi           2  0.223 0.11173   1.347  0.263
## Residuals   180 14.926 0.08292
##       region02  region08 striatum_ventral
## mean 0.2145690 0.2434223        0.1592058
## sd   0.2955092 0.2826002        0.2856290

3.2.3 Post-hoc Tests

##   Tukey multiple comparisons of means
##     95% family-wise confidence level
## 
## Fit: aov(formula = r ~ roi, data = data_tdomain_rois)
## 
## $roi
##                                  diff         lwr        upr     p adj
## region08-region02          0.02885330 -0.09437285 0.15207945 0.8448882
## striatum_ventral-region02 -0.05536319 -0.17858934 0.06786296 0.5390475
## striatum_ventral-region08 -0.08421649 -0.20744264 0.03900966 0.2418751

3.2.4 Visualization

3.2.5 Create function for t-tests, ANOVA

3.3 Valence

3.3.1 ANOVA

##              Df Sum Sq Mean Sq F value Pr(>F)
## roi           2   0.05 0.02490   0.399  0.672
## Residuals   180  11.24 0.06245
##         region02   region08 striatum_ventral
## mean -0.08846393 -0.1034032      -0.06341868
## sd    0.24809676  0.2577733       0.24364368

3.3.2 Post-hoc Tests

##   Tukey multiple comparisons of means
##     95% family-wise confidence level
## 
## Fit: aov(formula = r ~ roi, data = data_valence_rois)
## 
## $roi
##                                  diff         lwr        upr     p adj
## region08-region02         -0.01493927 -0.12187963 0.09200109 0.9417107
## striatum_ventral-region02  0.02504525 -0.08189511 0.13198561 0.8448285
## striatum_ventral-region08  0.03998452 -0.06695584 0.14692489 0.6513341

3.3.3 Visualization

3.4 Outcome

3.4.1 ANOVA

##              Df Sum Sq  Mean Sq F value Pr(>F)
## roi           2 0.0079 0.003935   0.437  0.647
## Residuals   180 1.6207 0.009004
##          region02     region08 striatum_ventral
## mean -0.008641344 -0.007030246      0.006005002
## sd    0.089935872  0.094721741      0.099751198

3.4.2 Post-hoc Tests

##   Tukey multiple comparisons of means
##     95% family-wise confidence level
## 
## Fit: aov(formula = r ~ roi, data = data_outcome_rois)
## 
## $roi
##                                  diff         lwr        upr     p adj
## region08-region02         0.001611098 -0.03899318 0.04221538 0.9951643
## striatum_ventral-region02 0.014646346 -0.02595793 0.05525063 0.6708993
## striatum_ventral-region08 0.013035248 -0.02756903 0.05363953 0.7287671

3.4.3 Visualization

3.5 Outcome Valence

3.5.1 ANOVA

##              Df Sum Sq Mean Sq F value Pr(>F)  
## roi           2  0.107 0.05339   2.504 0.0846 .
## Residuals   180  3.837 0.02132                 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##         region02    region08 striatum_ventral
## mean -0.01766416 -0.03709473       0.02102035
## sd    0.15169658  0.13782806       0.14815226

3.5.2 Post-hoc Tests

##   Tukey multiple comparisons of means
##     95% family-wise confidence level
## 
## Fit: aov(formula = r ~ roi, data = data_outcval_rois)
## 
## $roi
##                                  diff          lwr        upr     p adj
## region08-region02         -0.01943057 -0.081911508 0.04305036 0.7430735
## striatum_ventral-region02  0.03868451 -0.023796428 0.10116544 0.3111227
## striatum_ventral-region08  0.05811508 -0.004365854 0.12059601 0.0742090

3.5.3 Visualization

3.6 Outcome Valence Interaction

3.6.1 ANOVA

##              Df Sum Sq Mean Sq F value Pr(>F)
## roi           2  0.011 0.00539   0.136  0.873
## Residuals   180  7.141 0.03967
##        region02  region08 striatum_ventral
## mean 0.08384646 0.0894832       0.07112466
## sd   0.20692467 0.1841102       0.20568750

3.6.2 Post-hoc Tests

##   Tukey multiple comparisons of means
##     95% family-wise confidence level
## 
## Fit: aov(formula = r ~ roi, data = data_intract_rois)
## 
## $roi
##                                   diff         lwr        upr     p adj
## region08-region02          0.005636737 -0.07959765 0.09087113 0.9866253
## striatum_ventral-region02 -0.012721802 -0.09795619 0.07251259 0.9337446
## striatum_ventral-region08 -0.018358539 -0.10359293 0.06687585 0.8670500

3.6.3 Visualization

3.7 Combine one-sample t-test data

Fix up dataframe for pretty table

RSA results for a full and partial similarity analysis
ROI Mean df t-statistic p CI
14 CB Region 2 0.08 60 3.16 0.004 [ 0.03 , 0.14 ]
13 CB Region 8 0.09 60 3.80 0.001 [ 0.04 , 0.14 ]
15 Ventral Striatum 0.07 60 2.70 0.009 [ 0.02 , 0.12 ]
8 CB Region 2 -0.01 60 -0.75 0.640 [ -0.03 , 0.01 ]
7 CB Region 8 -0.01 60 -0.58 0.640 [ -0.03 , 0.02 ]
9 Ventral Striatum 0.01 60 0.47 0.640 [ -0.02 , 0.03 ]
11 CB Region 2 -0.02 60 -0.91 0.367 [ -0.06 , 0.02 ]
10 CB Region 8 -0.04 60 -2.10 0.119 [ -0.07 , 0 ]
12 Ventral Striatum 0.02 60 1.11 0.367 [ -0.02 , 0.06 ]
2 CB Region 2 0.21 60 5.67 0.000 [ 0.14 , 0.29 ]
1 CB Region 8 0.24 60 6.73 0.000 [ 0.17 , 0.32 ]
3 Ventral Striatum 0.16 60 4.35 0.000 [ 0.09 , 0.23 ]
5 CB Region 2 -0.09 60 -2.78 0.011 [ -0.15 , -0.02 ]
4 CB Region 8 -0.10 60 -3.13 0.008 [ -0.17 , -0.04 ]
6 Ventral Striatum -0.06 60 -2.03 0.046 [ -0.13 , 0 ]
17 CB Region 2 0.06 60 7.52 0.000 [ 0.04 , 0.08 ]
16 CB Region 8 0.06 60 5.16 0.000 [ 0.04 , 0.09 ]
18 Ventral Striatum 0.02 60 3.23 0.002 [ 0.01 , 0.04 ]
20 CB Region 2 0.07 60 9.00 0.000 [ 0.06 , 0.09 ]
19 CB Region 8 0.12 60 6.96 0.000 [ 0.08 , 0.15 ]
21 Ventral Striatum 0.02 60 3.65 0.001 [ 0.01 , 0.03 ]

4 Behavioral Analysis

Research Question: Does greater similarity for positive and negative wins relate to anxiety and depression?

Hypothesis 1: Greater similarity between positive and negative wins in the ventral striatum will be related to greater anxiety and less depression (Quarmley replication)

Hypothesis 2: Greater similarity between positive and negative wins in the cerebellum …

4.1 Hypothesis 1 testing

##                     Df Sum Sq Mean Sq F value Pr(>F)
## ch_totanx            1 0.0148 0.01479   0.570  0.458
## dep_group            1 0.0101 0.01012   0.390  0.539
## ch_totanx:dep_group  1 0.0044 0.00444   0.171  0.683
## Residuals           22 0.5708 0.02594
##             Df Sum Sq Mean Sq F value Pr(>F)
## ch_totanx    1 0.0148 0.01479   0.606  0.444
## Residuals   24 0.5853 0.02439

## [1] -0.04553422

4.2 Hypothesis 2

##                        Df Sum Sq  Mean Sq F value Pr(>F)
## ch_totanx               1 0.0148 0.014789   0.559  0.463
## ch_cdi_total            1 0.0009 0.000893   0.034  0.856
## ch_totanx:ch_cdi_total  1 0.0024 0.002442   0.092  0.764
## Residuals              22 0.5820 0.026454
##                        Df Sum Sq  Mean Sq F value Pr(>F)
## ch_totanx               1 0.0148 0.014789   0.559  0.463
## ch_cdi_total            1 0.0009 0.000893   0.034  0.856
## ch_totanx:ch_cdi_total  1 0.0024 0.002442   0.092  0.764
## Residuals              22 0.5820 0.026454

5 Age

## 
## Call:
## lm(formula = z ~ age, data = vs_vcr_social_fltr)
## 
## Residuals:
##       Min        1Q    Median        3Q       Max 
## -0.095566 -0.028568  0.001382  0.028446  0.096639 
## 
## Coefficients:
##              Estimate Std. Error t value Pr(>|t|)  
## (Intercept)  0.040064   0.022327   1.794   0.0779 .
## age         -0.001107   0.001261  -0.878   0.3836  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 0.04535 on 59 degrees of freedom
## Multiple R-squared:  0.01289,    Adjusted R-squared:  -0.003839 
## F-statistic: 0.7705 on 1 and 59 DF,  p-value: 0.3836
## 
## Call:
## lm(formula = z ~ age, data = cb_vcr_social_fltr)
## 
## Residuals:
##      Min       1Q   Median       3Q      Max 
## -0.25992 -0.09287 -0.02045  0.09230  0.56124 
## 
## Coefficients:
##             Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.065598   0.069411   0.945    0.348
## age         0.003257   0.003920   0.831    0.409
## 
## Residual standard error: 0.141 on 59 degrees of freedom
## Multiple R-squared:  0.01157,    Adjusted R-squared:  -0.005185 
## F-statistic: 0.6905 on 1 and 59 DF,  p-value: 0.4093